Implementing Cleaner Technologies in the
Printed Wiring Board Industry:
Making Holes Conductive
Design for the Environment Program
Economics, Exposure and Technology Division
Office of Pollution Prevention and Toxics
U.S. Environmental Protection Agency
Washington, DC 20460
This document was produced under grant #X-824617 from
EPA's Environmental Technology Initiative program.
-------
-------
ACKNOWLEDGMENTS
This report was prepared by Abt Associates Inc. for Microelectronics and Computer Technology
Corporation and The Institute for Interconnecting and Packaging Electronic Circuits as part of a
multi-stakeholder collaborative, Design for the Environment project. The EPA Project Officer
was Kathy Hart of the Design for the Environment Staff of the Office of Pollution Prevention and
Toxics. This report would not have been possible without the assistance of the technology
vendors and their customers who voluntarily participated in the interviews summarized in this
document. The project Core Group provided valuable guidance and feedback throughout the
preparation of the report. Core Group members included: Kathy Hart of U.S. EPA; John Lott of
DuPont Electronic Materials; Michael Kerr of Circuit Center Inc.; Jack Geibig, Lori Kincaid, and
Mary Swanson of University of Tennessee Center for Clean Products and Technology; Greg Pitts
of MCC; Christopher Rhodes of IPC; Gary Roper of H-R Industries, Inc.; John Sharp of Merix
Corp.; arid Ted Smith of Silicon Valley Toxics Coalition.
-------
-------
TABLE OF CONTENTS
PAGE
1. INTRODUCTION 1
2. CARBON METHOD . 4 3
Blackhole* (MacDennid, Inc.) .3
3. GRAPHITE METHODS ...* 11
Graphite 2000™ (Shipley Company) 11
Shadow™ (Electrocheimicals, Inc.) 15
4. PALLADIUM METHODS 19
ORGANIC-STABILIZED METHOD:
Neopact (Atotech U.S.A., Inc.) * * 21
TIN-STABILIZED METHODS:
Conductron DP (LeaRonal Inc.) 25
Crimson 1* (Shipley Company) .....28
Envision DPS™ (Enthone-OMI, Inc.) 33
HN504™ (Solution Technology Systems) 37
5. CONDUCTIVE POLYMER METHOD 41
Compact CP (Atotech U.S.A., Inc.) 41
6. OTHER ALTERNATIVE TECHNOLOGIES 47
7. LESSONS LEARNED 49
FACILITY INFORMATION APPENDIX A
-------
LIST OF FIGURES
FIGURE 1. TYPICAL CABBON PROCESS STEPS
PAGE
4
FIGURE 2. TYPICAL GRAPHITE PROCESS STEPS
FIGURE 3. TYPICAL PALLADIUM PROCESS STEPS [[[ 20
-------
1. INTRODUCTION
Direct metallization has been gaining attention in the printed wiring board (PWB) industry as an
environmentally preferable alternative to electroless copper. Even though many PWB
manufacturers are currently using these alternatives, there is still a lack of available information on
successfully implementing them. Some of the best sources of information about alternative
technologies for Making Holes Conductive (MHC) are those PWB manufacturers who have
actually installed and used the direct metallization systems under real-world operating conditions.
By sharing information, PWB manufacturers can benefit from others' experiences with the
relatively new technologies. This report details the specific experiences of these companies, along
with their recommendations for successful implementation.
This guide presents first-hand accounts of the problems, solutions, time, and effort involved in
implementing alternative MHC technologies. The information presented is based on telephone
interviews with PWB manufacturers currently using these technologies, manufacturers who have
used and discontinued these technologies, and the vendors of these alternative technologies.
When a respondent is directly cited, their comments appear in quotes. With the information from
these interviews, manufacturers considering a switch to an alternative technology can benefit from
the lessons learned by those who have already made the change.
Twenty PWB manufacturers and seven vendors were interviewed. Appendix A provides
background information on these facilities, such as annual production, types of boards produced,
highest aspect ratio normally run, and conveyorized (horizontal) or non-conveyorized (vertical)
configuration.
Carbon, graphite (two types), palladium (five types), and conductive polymer technologies are
discussed in this guide. Each section begins with a description of the technology, presents a flow
chart of the technology's typical process steps, and provides a summary of the interviews.
Introduction
-------
This document was developed as part of the Design for the Environment (DfE) Printed Wiring
Board Project. The DfE PWB Project is a voluntary, cooperative partnership which identifies and
assesses environmentally beneficial technologies and practices for the PWB industry. Project
partners include the U.S. Environmental Protection Agency (U.S. EPA), the printed wiring board
industry, Microelectronics and Computer Technology Corporation (MCC), the University of
Tennessee Center for Clean Products and Clean Technologies, the Institute for Interconnecting
and Packaging Electronic Circuits (IPC), and other stakeholders. The primary focus of the
project has been the evaluation of environmentally preferable MHC technologies. Quantitative
performance testing (both electrical and mechanical), risk characterizations, and cost analyses
were conducted on a comparative basis for several MHC technologies, including electroless
copper. The results of these analyses will be presented in a Cleaner Technologies Substitute
Assessment (CTSA) report.
Throughout the document, the facilities interviewed are not mentioned by name. Instead, each
company has been given a code, Facility A, Facility B, etc. It was the opinion of the project
participants that using the actual facility names might distract the reader from the information
presented on the technologies. However, the information in Appendix A will assist the reader in
understanding some of the circumstances governing production at each of the facilities
interviewed. This information includes the surface square feet produced annually, the types of
PWBs produced, and the highest aspect ratio run for each facility.
It should be noted that mention of trade
names in this report does not constitute
endorsement or recommendation for use.
Instead, the reader is encouraged to
contact the individual companies for more
information on their products.
Introduction
-------
2. CARBON METHOD
In carbon processes, a conductive layer of carbon black particles is deposited onto the substrate
surface and the through-holes. A pre-treatment conditioner solution removes oil and debris from
the substrate and creates a positive charge on the glass and epoxy walls of the vias. After
conditioning, the substrate is placed in a carbon black dispersion. A noncrystalline structure of
carbon black particles is adsorbed onto the positively charged surfaces, creating a conductive
layer coating the entire panel. A copper microetch then removes the carbon from the copper
surface while cleaning the surface for plating. Because the microetch does not attack the glass and
epoxy surfaces of the through-holes, a conductive carbon layer remains only on the through-hole
surfaces.
A typical carbon process has six chemical process steps (cleaner, carbon black, conditioner,
carbon black, microetch, and anti-tarnish) and two air knife/oven drying steps, as shown in Figure
1. The system is configured as an enclosed, conveyorized (horizontal) process. The specific
number and location of rinses and air knife stations depends on the type of product run at a
facility and the condition of the rinse water used.
Information is presented on the following carbon method:
*• Blackhole® (MacDermid, Inc.)
Blackhole® (MacDermid, Inc.)
Background
To date, approximately 135 Blackhole® systems, distributed by MacDermid, Inc., have been
installed worldwide. Blackhole® customers run a variety of substrates including Teflon®,
Carbon Method
-------
Cleaner
I
Carbon Black
Air Knife/Dry
Conditioner
Carbon Black
Air Knife/Dry
Microetch
Anti-Tarnish
Figure 1. Typical Carbon Process Steps
Carbon Method
4
-------
polyimide, and rigid flex, with holes as small as 0.008 inches in diameter. Most Blackhole®
customers run multi-layer boards. MacDermid has not identified any limitations in the types of
boards that can be run through the Blackhole® system.
REASONS WHY FACILITHS
SWITCHED TO BLACKHOLE®
»• Reduced cycle time
*• Hedaeed waste teeatmsat
*•
*" Wider process wHtdow
Implementation at Specific Facilities
For this implementation guide, two facilities in
the U.S. were interviewed about their
experiences with their Blackhole® systems.
Both facilities process primarily multi-layer
boards. Facilities A and B both run boards
with up to sixteen layers and boards with
aspect ratios of 8:1. Facility B has also
successfully processed boards with a 10:1
aspect ratio, and runs a wide range of
thicknesses from 0.001-inch thick flex to 0.250-inch thick back panels. When processing high
aspect ratio boards, Facility B runs them through the Blackhole® system twice, as "insurance,"
although the facility engineer thinks this step is probably unnecessary. Both of these facilities are
quick-turn shops, so reducing cycle time was the primary factor in their decisions to switch from
electroless copper to the Blackhole® system. Both, however, noted other potential benefits,
including reduced time and expenses for waste treatment system maintenance.
Experiences with the Blackhole® Process
Both facilities interviewed completed their Blackhole® installations within the last few months and
had very different experiences. Facility A installed Hollmuller equipment. After an installation
period of two to three weeks, the facility put product on the line and ran the system for one
month to qualify it. There were very few problems during the installation and debug period at
Facility A, other than some rollers that were redepositing carbon on the panel surface. The type
of roller used was changed and the problem was eliminated.
Carbon Method
-------
Facility A may have avoided other problems for two reasons. First, this facility uses deionized
water for the Blackhole® line. According to MacDermid, there have been problems at other
facilities where the water is very high in salts, which may contaminate the Blackhole® dispersion;
In these cases, the incoming water may need to be deionized. Second, the engineer at Facility A
noted that the capacity of their Blackhole® system far exceeds their throughput. This excess
capacity makes it easy for the facility to keep the process controls tight and within MacDermid's
parameters.
The installation at Facility B did not go as smoothly. The facility installed the first MacDermid
brand equipment manufactured in the U.S. Both "minor irritations" and "major problems"
occurred during the installation, all of which were equipment-related. MacDermid made all the
on-site modifications necessary to get the system working well, including changing the microetch
pumps, the cooling coils, the chiller, and the air knives. The engineer at Facility B believes that
the MacDermid equipment manufactured in Germany is superior, since manufacturers there have
had much more experience. Facility B's installation spanned several months, followed by one
month of running boards to qualify the system. Once all equipment-related problems were solved
(there were no issues with the chemistry), the facility was pleased with the system performance.
Facility B has made one equipment modification and one change in the process chemistry. The
facility found that propagation improved by running a lower microetch rate. The facility adjusts
the microetch daily to 35 to 40 microinches (win), instead of the 30 to 50 //in given in
MacDermid's specification sheet. The facility has also modified the first air knife section. The
system had holes in the lid for heat exhaust; however, the carbon solution was blowing off of the
boards and out of the holes in the cover. The facility mounted a piece of standard fiberglass door
screen to the lid, which let the heat out but kept the carbon in.
As a quick turn-around shop, Facility A deals primarily with engineers. Because of this
relationship, the facility did not have any problems obtaining customer acceptance of the new
system. The process engineer at Facility A noted that there may be additional hurdles to
Carbon Method
-------
acceptance for those companies that process a higher volume of boards. Facility B, also a quick
turn-around shop, had no problem gaining customer acceptance of the new technology. Facility B
invited its major customers (including NASA, the nuclear industry, and the military) to review the
data, look at the process, and evaluate the product. These two facilities advise others going
through this process to "let your customers tell you what they need to see" and then to supply it.
Comparisons to Electroless Copper
The two facilities interviewed experienced similar benefits of the Blackhole® system over
electroless copper. Both saw notable reductions in cycle time ("dramatically reduced" and
"significant decrease"). For example, the electroless copper line at Facility B took 1.5 hours to
get the first product through after running a load of dummies. After start-up, the line could
process approximately 60 panels per hour. Using the Blackhole® system, it takes only six minutes
to get the first panel through, dummies are not required, and product can be processed at an
average speed of 75 panels per hour.
Although quantitative data are not yet available on changes in board failure rates at either facility,
engineers from both facilities have noticed improvements. Facility A noted that the boards are
"far superior in terms of hole wall reliability." Facility B has seen increased capability in addition
to improved board quality. For example, Facility B's engineer noted that the facility can process
smaller holes with the Blackhole® system than with electroless copper. Also, when running
electroless, the Facility B operator had to visually scan every panel for problems. Because of the
improved quality of the Blackhole® system, this time-consuming step is no longer necessary.
Both facilities made changes to other parts of their process after the Blackhole® line was installed.
Facility A has benefited from a more consistent surface than on panels processed with electroless
copper. This has enabled the facility to eliminate one of the two acid cleaners from its pre-clean
line in the plating process. At Facility B, a downstream change was needed in electrolytic plating
operations. When using electroless copper, the facility would start with a low current density and
ramp up. With Blackhole®, this facility starts with a high current density for fifteen minutes to get
Carbon Method
-------
propagation through the hole, and then reduces the density to the lower level. Facility B was able
to eliminate the scrubbing process it previously used with the electroless copper line. This has
improved the facility's processing efficiency as the panels travel automatically from the
Blackhole* line into imaging. It should be noted that the Blackhole® line is a "no-scrub" system;
that is, scrubbing is not needed and should not be done after Blackhole® processing.
Although cost was not a primary motivation for installing the Blackhole® system, Facility A
estimates that its production cost per square foot has decreased with the new system. The cost
savings are the result of reduced chemical and labor costs. The labor savings come from a
reduction in the time required for lab testing and maintenance. The electroless system at Facility
A required from two to three hours per day of testing and maintenance, compared to two to three
hours per week for the Blackhole® line. Facility B has seen similar reductions. The facility now
spends about thirty minutes daily on lab analysis with the Blackhole® system, instead of over 2.5
hours per day with the old electroless copper system.
The two facilities have also realized cost savings through reduced maintenance requirements.
Weekly maintenance tasks include approximately two hours per week for chemistry changes and
four hours per week for equipment maintenance such as cleaning rollers and strainers, and
inspecting filters, nozzles, and air knives. As part of their system maintenance, the operator at
Facility B completes a 10-minute equipment check list and cleans the pinch rollers every morning.
Without cleaning, the carbon can get baked onto the pinch rollers over time, so the rollers are
removed daily and cleaned with water. MacDermid also stresses the importance of equipment
maintenance. The vendor recommends performing preventive maintenance for two to three hours
per week, including cleaning the carbon from the rollers and the nozzles twice a week.
Both facilities have also seen improvements in their waste treatment. Facility A has eliminated
batch treatment of chelated wastewater at 700 to 800 gallons a month. There was also a
"significant simplification" of waste treatment at Facility B. With the electroless copper line,
Facility B's wastewater discharge contained 2.5 parts per million (ppm) copper and "it was a job
Carbon Method
-------
to keep it there." With the Blackhole® system, the only treatment concern is the copper in the
microetch, which is less than 1 ppm with less treatment. Facility B also no longer has to treat the
manganese, several forms of copper, palladium, and chelated copper that were in the electroless
copper wastestream. Total water use at Facility A appears to be about the same for Blackhole® as
it was for its electroless system, whereas Facility B has seen a "considerable decrease" in its water
use.
"Dm *t sMmp on the eqmpm&nt*. .It
will end ttp c&stmg a, lot more for a
lot U$$ qmlity," %
-F&ctlityA
Keys to Success
Both the facilities and the supplier emphasized
the importance of quality equipment. The
process engineer at Facility A advises other
facilities, "Don't skimp on the equipment.
You'll end up with a sub-standard system that
will require all kinds of on-site modifications.
It will end up costing a lot more for a lot less
quality." Facility B experienced such a string
of on-site modifications. After several difficult months, the engineer at Facility B now considers
the system to be a "surprisingly pleasant experience." Having used and removed a palladium
system prior to Blackhole®, he advises facilities considering a change to talk to as many current
customers as possible and to run some product at a Blackhole® customer's facility. He believes
that Blackhole® has significant advantages over palladium processes, including a wider process
window and a shorter cycle time. According to the engineer interviewed, Blackhole® may have a
wider operating window than electroless.
Facility A also advises that facilities changing to direct metallization need to identify and
understand the current and anticipated problems in all parts of their production process. This is
because quality problems in other parts of the process can surface when a facility switches to
direct metallization. It may be that these problems always existed, but could not be detected until
direct metallization was installed. For example, it is important that problems in drilling or
Carbon Method
-------
desmear operations are corrected prior to installing the Blackhole® process. "Electroless copper
can be a band-aid over problems in other parts of the manufacturing process," according to
Facility A. His advice is to know where your problems lie and don't be "too quick to point the
finger at direct metallization."
MacDermid concurred with these observations, stressing the importance of working with the
vendor to evaluate the application before implementing any changes. Most vendors will help a
facility to determine if its line is suitable for direct metallization.
For more information on the Blackhole® process, contact Bill Sullivan of MacDermid, Inc. at 203-
575-5659.
Carbon Method
10
-------
3. GRAPHITE METHODS
Graphite methods disperse graphite (another form of carbon) onto the substrate surface. Similar
to the carbon method, a conditioner solution creates a positive charge on the substrate surface,
including the through-holes. Graphite particles are then adsorbed onto the exposed surfaces. In
contrast to the amorphous structure of the carbon black crystallites, graphite is a three-
dimensional, crystalline polymer. This crystalline structure creates a conductive layer covering
both the copper and the nonconductive surfaces of the outside layer and interconnects. A copper
microetch removes the unwanted graphite from the copper surfaces, leaving a conductive,
graphite layer on the glass and epoxy surfaces in the vias.
A typical graphite process has three or four chemical process steps (cleaner/conditioner, graphite,
fixer [optional], and microetch) and one air knife/oven drying step, as shown in Figure 2. The
number and location of rinses needed between process steps will vary by facility.
Information is presented on the following graphite methods:
+ Graphite 2000™ (Shipley Company)
*• Shadow™ (Electrochemicals, Inc.)
Graphite 2000™ (Shipley Company)
Background
The Graphite 2000™ process uses a patented shear pump to keep the graphite suspended in
solution. The Graphite 2000™ process is run on conveyorized (horizontal) equipment. Customers
Graphite Methods
11
-------
Cleaner/
Conditioner
Graphite
I
Fixer
(optional)
Air Knife/Dry
Microetch
Figure 2. Typical Graphite Process Steps
Graphite Methods
12
-------
using this process run predominantly double-sided boards (approximately 65 to 90 percent
double-sided and 10 to 35 percent multi-layer). Most of the customers run between 2,000 and
5,000 surface square feet per day. According to the vendor, the Graphite 2000™ system is limited
to boards that are 0.125 inches thick or less, holes that are 0.013 inches in diameter or greater,
and aspect ratios of 8:1 or less.
Implementation at Specific Facilities
Two facilities (Facilities C and D) that have
successfully implemented the Graphite 2000™
process were interviewed for this report.
Primary motivations for switching to the
Graphite 2000™ system in both cases included
the elimination of formaldehyde, hydrazine,
and cyanide; reduced operating costs; and
improved worker safety. The facilities chose
the Graphite 2000™ system for several
reasons: the vendor had a strong reputation in
the industry, a good relationship had been established with the vendor, and both facilities were
beta sites for the technology.
SWITCHED TO GfcAMf i £000
* BlWaatioft of formaldbfeyde
* Lower operating costs
»Improved worker safety
» Less water consumption
* Reduced cycle time
Facility C was the alpha-beta site; it was the first facility to install the Graphite 2000™ system.
Facility C took three months to install and debug the Graphite 2000™ system; Facility D took six
weeks. Both facilities installed new equipment from Finishing Services Limited (FSL), one of the
vendors recommended by Shipley. To reduce water usage, Facility C installed a chiller to provide
a closed-loop cooling system for the conveyorized unit. Facility D modified its equipment by
removing the scrubbing unit and adding an anti-tarnish module and a high-pressure (125 pounds
per square inch) water blast at the end of the graphite line. According to Shipley, equipment
installation requires a week, chemistry evaluation takes another week, and then the system needs a
trial month before a facility can go to full production on all substrates and work types.
Graphite Methods
13
-------
Experiences with the Graphite 2000™ Process
During debugging, equipment problems outnumbered chemistry problems. For example, Facility
C experienced problems with plugged nozzles that caused weak coverage on the board. By
replacing the spray manifolds with fluid wedges, the facility improved the coverage and solved the
problem. Occasionally, the squeegee rollers do not remove enough water from the board (after
the graphite tank) during full production. Daily preventive maintenance helps minimize equipment
problems. "Sometimes," reported Facility C, "small holes can be an issue. In these cases, we run
the boards through three times." The facility believes that Shipley is improving the process so
that smaller hole sizes can be run. Approximately 97 percent of Facility C's customers accepted
the new technology immediately; the remaining 3 percent needed more data, testing, and in-house
inspections before accepting it. Customers of Facility D had no problems with the new
technology.
With its previous electroless copper line, Facility D contracted out its multi-layer production so
that the facility would not need a permanganate desmear operation in-house. When Facility D
installed its new Graphite 2000™ line, the facility decided to continue to contract out its multi-
layer production
Comparisons to Electroless Copper
Facility C spends more time on equipment maintenance for the Graphite 2000™ process than for
the electroless copper process, but less time on lab analysis. Facility D has not experienced any
major changes in time spent on maintenance or lab analysis. Both facilities report reduced cycle
time and water usage with the new system.
Keys to Success , -.. '.
Both the vendor and the manufacturers thought that commitment and dedication at all levels, from
management down to the line operators, is vital for the successful implementation of the Graphite
2000™ system. According to Facility D, "The selection of high-quality equipment, and its daily
maintenance, are extremely important. Also, management must be patient. It takes four to six
Graphite Methods
14
-------
weeks to get everything working smoothly."
Shipley believes that a company needs to
view the Graphite 2000™ process as part of
the bigger picture. "Often, changes need to
be made upstream or downstream to optimize
the graphite process. A company needs a
willingness and commitment to change their
process and to maintain better control of the
process." For example, a facility switching
to the Graphite 2000™ process might need to adjust and optimize the process window in the
electrolytic plating step downstream from the graphite step.
For more information on the Graphite 2000™ system, contact Hal Thrasher of Shipley Company
at 508-229-7594.
"The selection &f high-quality
equipment, anditsdaity
maintenance, am extremely
important,n
-Facility £)
Shadow™ (Electrochemicals, Inc.)
Background
The Shadow™ process uses a patented binder system in the graphite mixture to promote hole wall
adhesion and colloid stability. The process also includes a fixer step immediately following the
graphite bath; the patented fixer promotes a uniform graphite coating of the hole wall. Almost all
Shadow™ systems are conveyorized (horizontal). The Shadow™ process has successfully run
multi-layer boards and exotic substrates (e.g., Teflon®, rigid flex). According to the vendor, the
limitations of the system are related to the quality of the incoming boards; drilling quality is
especially important. According to one facility interviewed (Facility E), boards that are thicker
than 0.093 inches are run at a slower conveyor speed, and Teflon® boards go through two passes
at the slower conveyor speed.
Graphite Methods
15
-------
Implementation at Specific Facilities
Three facilities (Facilities E, F, and G) were
interviewed for this report—two that have
successfully implemented Shadow™ and one
that has switched back to electroless copper.
Primary motivations for switching to the new
process included the elimination of
formaldehyde, reduced cycle time, and
decreased water usage. One facility
interviewed chose the Shadow™ system
because it was the most affordable conveyorized system available at the time (March 1995). A
second facility interviewed helped the supplier develop and test the Shadow™ system. Installation
of the system took approximately two to five days (not counting delays due to missing equipment
parts), and the debugging period ranged from one to two months. All three facilities purchased
new, conveyorized (horizontal) equipment for the system.
REASONS WHY FACILITIES
SWITCHED TO SHADOW™
> Elimination of formaldehyde
> Reduced cycle time
»• Less water consumption
* Affordable for a conveyorized system
Experiences with the Shadow™ Process
All the facilities encountered some problems during debugging and/or full operation, most of
which were equipment-related. Facility E now experiences only occasional mechanical problems
during full production; "there are always little problems." For example, if one roller is out of
place, it creates dragout which could contaminate the other tanks. Another problem occurs when
the squeegee rollers sometimes develop hard spots where solids collect. "We are thinking of
having a second set of rollers immersed at all times, so that changeover is more efficient," the
facility reported. Most equipment problems can be minimized by aggressive preventive
maintenance.
During debugging, Facility F found that graphite left on the board surface due to excessive
dragout created drying problems. To reduce dragout from the graphite tank, the facility increased
the tension of the squeegee rollers. This facility also had graphite build-up at the "knee" of the
Graphite Methods
16
-------
holes during drying. To solve this problem, the facility installed a bigger blower motor in the
dryer to create sufficient air flow through the holes. Facilities E and F have not had any issues
with customers accepting the Shadow™ process.
Facility G switched back to electroless copper after running the Shadow™ line for less than one
year. During debugging, Facility G switched from a sulfurie peroxide microetch to a sodium
persulfate microetch, which increased the copper discharge concentration. In addition, the facility
had problems maintaining the agitation needed to keep the graphite in suspension. This resulted
in unexpected sludge generation and plugged nozzles. During full production, the facility's
customers found that solder joints would fail during circuit board assembly.
According to the vendor, Electrochemicals, Inc., almost all of the companies that have pulled out
of the Shadow™ process had problems with equipment. The vendor stated that it is critical for
manufacturers to follow the vendor's equipment recommendations.
Comparisons to Electroless Copper
Facility F spent less time maintaining the Shadow™ line than maintaining their previous electroless
system, and spent a lot less time on lab analysis. Facility F reduced cycle time from 90 minutes to
approximately 10 to 15 minutes, while significantly reducing water consumption. Chelated
copper was eliminated from the wastestream, and the copper concentration in the discharged
wastewater was reduced.
Keys to Success
Electrochemicals, Inc. emphasizes the
importance of quality production practices.
"If a facility has quality problems with
electroless copper, it will still have those
problems with direct metallization." The
vendor believes it is important that facilities
"If £i facility has quality problems
with electroless copper, it mil still
have those problems with direct
metallization*"
-Electrochemical^ Inc.
Graphite Methods
17
-------
follow the vendor's recommendations for equipment purchases. The facilities report that daily
chemical analysis, preventive equipment maintenance, and a commitment to eliminating
formaldehyde are necessary to successfully implement the Shadow™ process.
For more information on the Shadow™ system, contact John Myers of Electrochemicals, Inc. at
612-479-2008.
Graphite Methods
18
-------
4. PALLADIUM METHODS
Palladium systems use palladium particles to catalyze nonconductive surfaces of the through-
holes. Palladium particles tend to agglomerate (cluster) unless they are stabilized through the
formation of a colloid, which surrounds the individual palladium particles with a protective layer.
The two main categories of stabilizers are organic polymer and tin.
Initially, a conditioner solution creates a positive charge on the substrate surface. For organic-
polymer/palladium colloids, a predip solution conditions the surfaces of the vias with a polymer
film that acts as an adhesion promoter for the colloids. When the substrate is introduced to
colloidal suspension, the tin/palladium colloids adsorb onto the slightly charged surfaces, and the
organic-polymer/palladium colloids adsorb onto the film-covered through-hole walls. The
adsorbed colloidal particles form a nonconductive coating on the through-hole walls. The
substrates are then placed in an accelerator solution (for tin) or a postdip solution (for organic
polymer) which removes the stabilizers, exposing a conductive layer of palladium particles in the
through-holes.
A typical palladium process has six chemical process steps (cleaner/conditioner, microetch,
predip, catalyst/conductor, accelerator/postdip, and acid dip), as shown in Figure 3.
Information is presented on the following palladium methods:
Organic-stabilized method:
>• Neopact (AtotechU.S.A., Inc.)
Palladium Methods
19
-------
Cleaner/
Conditioner
Microetch
Predip
I
Catalyst/
Conductor
I
Accelerator/
Postdip
Acid Dip
Figure 3. Typical Palladium Process Steps
Palladium Methods
20
-------
Tin-stabilized methods:
> Conductron DP (LeaRonal Inc.)
> Crimson 1® (Shipley Company)
*• Envision DPS™ (Enthone-OMI, Inc.)
* HN504™ (Solution Technology Systems)
Neopact (Atotech U.S.A., Inc.)
Background
Neopact is an organic-stabilized palladium-based technology available in both non-conveyorized
(vertical) and conveyorized (horizontal) configurations. The Neopact system is quite versatile,
working on a wide variety of substrates, boards with many layers, and boards with very small
holes. Many Neopact customers produce complex boards, typically with ten or more layers.
Although electroless copper is likely to be more effective than direct plate for extremely thick
boards, the difference in quality is narrowing with experience, according to the vendor, Atotech
U.S.A., Inc. Hole diameter sizes of 0.008, 0.010, 0.013 inches, and some as small as 0.006 inches
have been through the qualification process. Neopact users work with a wide variety of
substrates, including FR-4, Teflon®, polyimide, acrylic flex, non-acrylic flex, and epoxy.
EEASOHS WE? A FACILITY
SWITCHED TO
Implementation at Specific Facilities
Two facilities (Facilities H and I) that have
successfully implemented the Neopact process
were interviewed for this report. Their
primary motivations for switching; to the new
process included the elimination of
formaldehyde, lower costs for labor and
support materials, and decreased water use.
Facility H chose the Neopact process because it was capable of running the wide variety of
* EBisiiaatiott of formaldehyde
>• Reduced labor and material costs
>• Decreased water ase
* Ability to run a variety of substrates
Palladium Methods
21
-------
substrates that the facility uses. In addition, some systems offered by other vendors utilize a
permanganate desmear, which only works on epoxy. The use of permanganate would not have
been compatible with Facility H's plasma desmear operation. Facility I switched to the Neopact
system in January 1996 after nine months of experiencing problems with voids using a palladium
process supplied by a different vendor.
Both facilities used an existing, computer-controlled, non-conveyorized (vertical) rack system for
the Neopact process. Installation at Facility H took four days, and there were no unexpected
expenses. Facility H's only equipment modification was the addition of a heating and cooling coil
to one of the process tanks. This facility first ran a prototype system, which underwent a
two-week debugging process followed by eight months of end-user qualification testing. The
company had to qualify the new process to meet military qualifications and other customer
requirements. At that point, the Neopact system was incorporated into the main production line.
Installation at Facility I took approximately two to three days. During this period, however, the
facility discovered additional equipment needs that were not anticipated. These unexpected
expenses for line set-up included heating coils on one tank and a flash-plate step to allow for void
detection.
The vendor, Atotech U.S.A., Inc., noted that when a Neopact system is installed in a
conveyorized (horizontal) configuration, the debugging timeline is much longer due to the
complex nature of the installation and start-up of automated equipment. It usually takes three to
five months to install and debug the chemistry and equipment. The vendor strongly recommends
using its own equipment (sold under the trade name Uniplate) to minimize debugging and control
problems.
Palladium Methods
22
-------
Experiences -with the Neopact Process
In Atotech's experience, "facilities retrofitting existing tanks for the Neopact system usually go
through a phase of four to six weeks during which a facility discovers unique qualities of its
process that require adjustments" (e.g., analytical frequency, dumping schedules, and interactions
with other equipment). This was true for Facility H, although its debugging period was shorter.
Other than minor fine-tuning, Facility H did not encounter any problems during debugging.
After the Neopact system was in full production at Facility H, problems began to occur with the
oxidation-reduction potential (ORP) controller on the palladium bath. Over time, the controller
and probe failed, yet the operator could not immediately detect its failure. The facility eventually
replaced the controller and probe with a more reliable one.
In contrast to Facility H, Facility I experienced problems with voids during the debugging
process. On the advice of the vendor, the facility solved the problem by adding an extra step to
the process (a "wetter" step), which required an additional tank and chemistry. A recycling pump
was also added to allow for better circulation of the chemistry through the holes. At Facility I, it
took about three months to get the system working properly. With the system fully operation,
Facility I periodically experiences voids on all types of boards. When this occurs, the facility
works with the local Atotech service representative to adjust the process chemistry. Inner and
outer layer separation on multi-layer boards is another problem that occasionally happens, the
solution to which the company is still investigating.
According to Atotech, voids have not been a commonly cited problem with the Neopact system.
To inspect through-holes for voids, customers use the backlighting technique after flash or panel
plating. If the customer pattern plates after direct plating, a backlight test coupon is used along
with a microsection of the finished product. It is not possible to perform backlight inspection of
the product unless it has been flash-plated. Regardless of the effective methods of inspecting for
voids, some PWB manufacturers may have customer specifications (e.g., military) to use the
solder float test followed by cross-sectioning. This is the case for Facility H. Neither company
Palladium Methods
23
-------
had significant problems with customer acceptance, although Facility H noted that some
companies may be reluctant to accept the process because it is new.
Comparisons to Electroless Copper
Facility H found that cycle time and labor time required for preventive maintenance on the
Neopact system are roughly the same as for electroless copper. However, the labor time
necessary for process control has decreased by 50%, and the time spent on lab analysis has also
been reduced. In addition, board quality is reported to be superior, and the board failure rate has
been halved. Facility H noted no differences in copper discharge, sludge generation, or water
usage. However, chelated wastestreams are not generated by the Neopact system.
The switch to the Neopact process at Facility I has resulted in a cycle time 50% faster than for
electroless copper. The facility notes no differences in board failure rate and the amount of time
spent on maintenance. Some changes in waste treatment have been required as a result of
implementing the new process. Palladium-containing wastewater cannot be treated in the
facility's resin-based treatment system, so it must be shipped off-site. On the other hand, some
process wastes do not require treatment at all before .discharge. With regard to environmental
impact, Facility I has achieved reduced copper discharge, and savings in water usage, but sludge
volume remains unchanged.
Keys to Success
Facility H emphasizes that close support from
the supplier during start-up is critical to the
success of this technology. PWB
manufacturers who choose to implement this
technology should request that the vendor
supply a technician who has substantial
experience setting up the system to work in
the facility during the start-up period. Facility
Atotech LLSiii, Inc. emphasized that
$ t W^ J MJ\
management must support alt phases
of the implementation, from
•> ... j •'••s- < if
installation to debugging to full
production.
Palladium Methods
24
-------
H attributes its success to a thorough evaluation of the process (e.g., plating distribution, post-
separation resistance, chemical usage) prior to full production in order to facilitate customer
approval.
For companies to be successful with the Neopact system, Facility I stresses that "training is very
important because you are working with very sensitive chemistry." Operators must maintain
process baths according to vendor specifications. "You can't stretch it when it's time to change
[the chemistry] or else you'll have problems." Facility I also considers flash plating (to facilitate
void detection) a necessary step "in order to have 100% confidence" that proper plating in
through-holes is achieved.
Atotech emphasized that management needs to make a firm commitment to the alternative
technology. Management must support all phases of the implementation, from installation to
debugging to full production.
For more information on the Neopact system, contact Mike Boyle of Atotech U.S.A., Inc. at 803-
817-3561.
Conduction DP (LeaRonal Inc.)
Background
The Conductron DP process accelerates the tin from the tin/palladium colloid, and at the same
time reduces the copper back onto the palladium. The resulting layer of conductive
palladium/copper is electroplated with copper. The Conductron process can be run for both non-
conveyorized (vertical) systems (Conductron DP) and conveyorized (horizontal) systems
(Conductron DP-H). According to the vendor, LeaRonal Inc., there are no substrate limitations,
and a maximum aspect ratio of 26:1 has been run on a conveyorized (horizontal) system. Most of
the facilities that run the Conductron process produce less than 500,000 surface square feet per
year, but some large facilities have successfully installed the system as well.
Palladium Methods
25
-------
REASONS WHY FACILITIES
SWITCHED TO CONPUCTRON DP
operating costs
+ Less water use
i
» Quicker throughput
* Increased worker safety
* Ease of waste treatment
Implementation at Specific Facilities
Two facilities (Facilities J and K) that have
successfully implemented the Conductron
system were interviewed for this report.
Primary motivations for switching to the new
process included lower operating costs,
reduced water use, quicker throughput,
increased worker safety, and ease of waste
treatment. One facility chose LeaRonaFs
Conductron system because the vendor had a
lot of experience in the industry, the initial test
results for Conductron were better than those for carbon and graphite systems, and the palladium
technology is similar to that of electroless copper, making it easier to sell to customers.
Installation and debugging of the non-conveyorized (vertical) system at Facility J took
approximately six months. Facility J did not purchase any new equipment and had no unexpected
expenses during this time. Other facilities, according to LeaRonal, had typical equipment issues
with the conveyorized systems during debugging. "With these complex systems, there is bound to
be a problem with pumps, wiring, etc." Also, tap water contaminated baths at some facilities. In
these cases, the facilities switched to deionized water.
Experiences with the Conductron Process
Facility J stated that "any printed wiring board manufacturer will experience intermittent
problems" with plating through-holes. During full operation, the facility occasionally experiences
variations in bath temperatures and contaminated chemistries caused by human error. The facility
encounters about the same number of problems as it did with the previous electroless copper line.
One customer—a military account—decided to take its business elsewhere, because it wanted a
technology backed by many years of test data on performance. Nevertheless, Facility J meets
military qualifications and can run Teflon®, FR-4, and polyimide substrates.
Palladium Methods
26
-------
Facility K primarily uses an electroless copper line, but recently installed a conveyorized
(horizontal) Conductron DP-H line to help with smaller orders that require quick turn-around
time (e.g., prototype boards). The facility encountered some problems during debugging and
testing of the Conductron line. There was excess dragout due to the squeegee rollers. This was
solved by resurfacing the rollers. While there was no problem with epoxy coverage, there was
inconsistent coverage for glass surfaces in the through-holes. Tin was also oxidizing out because
of poor machine design. The equipment would aerate the solution, resulting in tin precipitating
out. There were also problems with liquid level controls. To solve these problems, Facility K is
planning to install a new conveyorized Conductron system. The new Conductron line is not
meant to replace the existing electroless line.
Comparisons to Electroless Copper
Facility J spends more operator time on bath maintenance and lab analysis with the Conductron
system than it did with electroless copper, while Facility K predicts its new system will need less
time for analysis and chemistry bath maintenance. Facility J's Conductron line has a cycle time
that is 60 to 75 percent faster than the previous electroless line; Facility K predicts the
Conductron line's cycle time will be approximately 65 percent faster than electroless copper. The
Conduetron systems at both facilities generate less sludge and less copper waste.
Keys to Success
Facility J believes it is very important to have
excellent, well-established vendor support to
successfully implement a new direct
metallization system. The facility believes that
LeaRonal has provided good technical
support and is "very knowledgeable about the
chemistry." Also, the facility emphasized the
need for line operators who are willing to
learn about the new technology. Facility K stated that the equipment is very important, and that
Facility J believes it is very important
to have excellent^ well-established
vendor support to successfully
implement a new direct metallization
system.
Palladium Methods
27
-------
the vendor's equipment recommendations should be followed. The Conductron system,
according to both facilities, is "not as forgiving as electroless copper and requires more operator
attention."
For more information on the Conductron DP system, contact David Schram of LeaRonal Inc. at
516-868-8800.
Crimson 1® (Shipley Company)
Background
The Crimson 1® system is a tin-stabilized palladium process. Crimson 1® differs from other tin-
stabilized palladium processes in that it uses a sulfide step to stabilize the surface of the
accelerated substrate. After the sulfidization of the palladium sites, some sulfide adsorbs to the
exposed copper of the inner layers. This coating tints the boards a crimson color (and thus gives
the technology its name). A microetch step then removes the adsorbed sulfide from the
interconnects. The final step of the process, a high-pressure water rinse, removes any remaining
microetch from the board surface.
According to Shipley, 30 of 70 Crimson 1® facilities use a conveyorized (horizontal)
configuration. In a non-conveyorized (vertical) orientation, Shipley recommends that only
double-sided boards be processed, but a conveyorized (horizontal) system can process both multi-
layer and double-sided boards. This difference is due to the difference in the fluid dynamics. In a
horizontal (conveyorized) system, there is more control of the flow of the solution going through
the holes. This increased control is important in successfully running multi-layer boards. Most
new systems are conveyorized (horizontal) lines, according to the vendor, "now that more shops
are doing multi-layers." The typical Crimson 1® customer manufactures 60 percent multi-layered
boards in volumes that range from 500 to 2,000 panels per day, with hole diameters of 0.008
inches and larger. The vendor nptes, however, that 20 of these facilities process many more per
Palladium Methods
28
-------
day (for example, one processes 40,000 square feet per day of primarily double-sided boards,
while another handles 30,000 square feet per day, 90 percent of which are multi-layer).
According to the vendor, the current standard product mixes are being used successfully with the
Crimson 1® process, customers are successfully running exotic substrates through the system,
and, in some cases, aspect ratios as high as 20:1 have been plated successfully.
Implementation at Specific Facilities
Three facilities (Facilities L, M, and N) that
have successfully implemented the Crimson 1®
process were interviewed. All three facilities
run the process horizontally, with production
volumes ranging from 1.44 to 9.6 million
surface square feet per year. Their primary
motives for switching to Crimson 1® from
electroless copper were decreases in chemistry
costs, cycle time, water consumption,
maintenance, and waste treatment costs, as
well as improved worker health and safety. Debugging required the longest time at Facility M,
the first U.S. manufacturer to use the Crimson 1® process for multi-layer boards. This line, which
was installed in the summer of 1993, took one month for physical installation and nine months for
debugging. Facility L also installed its line that year, taking six months due to the large size of the
line. Debugging required four months. Facility N installed its Crimson 1® line in only two weeks
in June 1996, with debugging requiring three months.
All three facilities had researched or experimented with other technologies before adopting the
Crimson 1® process. One facility experimented with carbon technology, and another facility
experimented with graphite technology, but both experienced defect problems and overall process
sensitivity. With the Crimson 1® system, however, one facility said "we couldn't get it to fail."
Another facility remarked that "the Crimson process appeared to be more robust than other direct
REASONS WHY FACILITIES
SWITCHED TO CRIMSON 1*
*• Decreased chemistry costs
> Faster cycle time
»- Less water consumption
> Lower waste treatment costs
•* Improved worker health and safety
Palladium Methods
29
-------
metallization systems." The third facility selected Crimson 1® based on its own research on the
costs and voiding frequencies of alternative technologies.
Experiences with the Crimson 1® Process
Since Facility M's original persulfate microetch also etched the palladium, the facility switched to
a peroxide microetch. This change necessitated other downstream process changes. First,
different waste treatment chemicals were needed due to peroxide gassing; they now use sodium
metabisulfite to help suppress gas formation. Also, the sodium peroxide evaporated and needed
refreshing if the Crimson system was not in use. Lastly, Facility M noted that the sulfide waste
from the Crimson 1® process requires special handling~the waste should be added directly to the
batch treater to obviate hydrogen sulfide formation. The other two facilities interviewed also
made adjustments to the process once it was operational.
All facilities experienced some minor process defects. Facility M reported negative etchback with
Crimson 1* that was "noticeable" but within customer tolerances. This facility noted that military
or three-point connections (and thus military-specification boards) are not a possibility due to
etchback. Facility N experienced slight hole-wall separation. Facility L found that microetch
must be maintained within proper limits to insure consistent film removal. In addition, Facility L
found that a board scrubber was necessary to remove a thin film layer from the board surface
before sending boards to the lamination room. According to Facility N, Shipley now recommends
board scrubbing to manufacturers installing their lines.
Facility N had some conveyor jamming problems when running thin-core materials (e.g., 0.006
inches). Jams would occur when improperly adjusted water pressures knocked panels out of
clips. These problems were eliminated after spray pressures were adjusted and stronger clips
were added.
Manufacturers did not report major hole size or board thickness limitations with the Crimson 1®
technology. Facility N experienced some failures with hole diameters of 0.008 inches, but solved
Palladium Methods
30
-------
the problem by experimenting with operating parameters. Newer versions of the Crimson 1®
system have improved filtration and spray bar configurations, according to Facility M. A Florida
facility with a Crimson 1® line is reportedly plating hole diameters as small as 0.008 inches and up
to a 12:1 aspect ratio on "every material you can think of." According to Facility M, "Crimson 1®
can do everything, once properly configured."
Comparisons to Electroless Copper
All three facilities increased production throughput with the Crimson 1® system. Facility M
switched from a "typical electroless dip-and-dunk batch system" to the conveyorized (horizontal)
Crimson 1® process and saw large throughput increases. Facilities L and N doubled and tripled
their productivity by switching from electroless copper to the Crimson 1® system.
Manufacturers saw void frequency either decrease or stay the same. For Facility M, the Crimson
1® process "reduced voids by 90 percent, at least." The manufacturer at Facility M "used to tilt
boards, bang them, vibrate them, but still had a problem with voids" due to hydrogen bubbles in
the holes of the vertically-hanging boards. Facility N noted that the "failure rate is one-third of
electroless" and "more a function of drilling and drill debris than a failure of Crimson."
Time required for lab analysis decreased at all facilities. Facility M now conducts one analysis a
shift (eight hours) instead of once an hour. Facilities L and M report that continuous monitoring
is now not necessary and fewer parameters need to be analyzed overall.
Changes in maintenance requirements varied by facility. Facility M noted "We used to have to
use a colorimeter to calibrate baths and pumps and were constantly fighting the metering pumps,
but now we don't have to worry about it at all." Facilities M and N caution that "there is a lot of
preventive maintenance on the [Crimson 1®] line because it is so complicated." At Facility N,
time spent on maintenance increased because of an extensive preventive maintenance schedule.
Palladium Methods
31
-------
Facilities M and N saw water usage decrease. Facility M's water usage with the electroless
system was 10 to 14 gallons per minute. Water use is 3 to 4 gallons per minute with Crimson 1®
because the "rinsing tanks on the Crimson line have a far lower rating than the electroless."
Facility L did not track water usage.
For Facility M, the switch from electroless copper to the Crimson 1® process did not appreciably
change the concentration of copper in facility wastewater or the amount of sludge generated.
However, Facility N saw a decrease in both. Facility L did not know if these factors had changed.
All facilities reported lower air emissions (such as formaldehyde).
All three facilities reporting saving money with the Crimson 1® system. Facility N reported saving
50 to 60 percent overall compared to the costs of an electroless system. Facility M stated that the
Crimson 1® process itself cost the same as electroless, but savings were possible in other areas
such waste treatment, chemical maintenance, and lab analysis. This facility reported saving at
least 30 percent overall. Facility L reported saving roughly 50 percent on chemical costs, but did
not track other cost changes.
Keys to Success
Preventive maintenance is crucial to the
success of a Crimson 1® line, according to all
three manufacturers and the vendor. Facility
M recommended paying special attention to
keeping the fluid delivery systems free of
clogs and debris. Other issues included
profiling thin-core transport through the
system and avoiding variations in water
pressure (which can burn out pumps).
"You need to took at how the
manufacturing process overall wm
change. A total process mentality is
crucial."
^Shipley Company
Facilities and the vendor also recommended re-evaluating the entire process when adopting the
Palladium Methods
32
-------
Crimson 1® system. According to the vendor, "you can't just drop out electroless out and drop in
'Crimson.' You need to look at how the manufacturing process overall will change. A total
process mentality is crucial."
For more information on the Crimson 1® system, contact Hal Thrasher of Shipley Company at
508-229-7594.
Envision DPS™ (Enthone-OMI, Inc.)
Background
The Envision DPS™ method deposits a palladium-tin colloid on the hole during the activation
step. A highly alkaline (pH of >12) copper-containing solution at an elevated temperature is used
to substitute copper for tin through disproportionation (US patent 5,376,248). Electroplating
takes place on the resulting palladium-tin-copper film.
According to the vendor, Enthone-OMI Inc., all but one of the 25 manufacturers using the
Envision DPS™ operate it vertically, many in existing electroless copper equipment. Envision
DPS™ customers process a variety of board types and dielectric materials ranging from double-
sided FR-4 material to multi-layers, Teflon® and high Tg dielectrics.
Hole diameters of 0.018 inches in multi-layers which are 0.125 inches thick are successfully
processed using non-conveyorized (vertical) process configuration. The operation of the Envision
DPS™ process in conveyorized (horizontal) configuration improves solution exchange and
increases the operating window. Smaller hole sizes and thicker boards are possible.
Palladium Methods
33
-------
Implementation at Specific Facilities
Three facilities (Facilities O, P, and Q) that
have successfully implemented the Envision
DPS™ process were interviewed for this
report. Their motivating factors for switching
to the process from electroless copper were
environmental compliance, decreased waste
treatment and disposal costs, improved
throughput, and decreased use of toxic
chemicals (e.g., formaldehyde and cyanide).
REASONS WHY FACILITIES
SWITCHED TO ENVISION DPS™
*• Decreased waste treatment
» Better hole wall integrity
* Increased throughput
*• Less use of toxic chemicals
+ Compatible with electroless equipment
Facilities O and Q explored carbon and graphite technologies (one facility tested them side-by-
side) before adopting the Envision DPS™ system. These facilities mentioned wider operating
parameters and improved hole wall integrity as the primary reasons they chose Envision DPS™.
Facility Q also explored a different palladium system, but was discouraged by the up-front
investment required to implement the technology.
Installation and debug time ranged from one day for the retrofit of an old line to one month for
installation of an entirely new line. All three facilities use the non-conveyorized (vertical) process
for all of their board production. Facility Q noted that the Envision DPS™ line was "very
compatible" with the old electroless tanks, line set-up, and process. All three manufacturers
reported that existing tanks from an electroless copper line were used to some degree, minimizing
the need for new equipment purchases and extensive operator training. Additional expenses for
line set-up were minor, including an extra filter pump in the catalyst bath, multi-meters, heating
element adjustments, and modified pump capabilities.
Palladium Methods
34
-------
Experiences with the Envision DPS™ Process
All three facilities encountered minor problems during debugging. At Facility O, hole wall
adhesion problems on double sided PWBs were traced back to the drilling step. This
manufacturer changed drilling parameters1 and installed a water blast in their deburring operation
to remove loose hole debris. This resulted in improved hole wall quality and eliminated hole wall
adhesion failures.
Facility P noted that very large hole sizes tended to void slightly more often than would be
expected with an electroless bath. These voids were not seen as a function of poor bath
chemistry, but, according to the manufacturer, may have been a result of smears from the drilling
process. Manufacturers did not encounter other hole size, board type, or board thickness
limitations for Envision DPS™.
Reworked Envision DPS™ panels needed to be treated with slightly more care than reworked
electroless panels at Facility Q. When the stripper bath was used to remove dry film, the bath also
occasionally removed the palladium from the holes. The manufacturer reran the panels when this
occurred.
The vendor cautioned that the alkaline nature of some of the steps in the Envision DPS™ process
may cause problems with alkaline-sensitive adhesives. These issues are similar to those that
facilities may experience with electroless copper, but with more pronounced adhesive swelling.
According to the vendor, adhesive swelling issues are not a problem in the conveyorized
(horizontal) mode due to reduced contact time with the alkaline solutions.
This manufacturer changed drill bits, cut the number of hits per drill bit from 2,500 to 1,000, and switched
from four panel to three panel stacks.
Palladium Methods
35
-------
Comparisons to Electroless Copper
All three manufacturers found the Envision DPS™ system more cost-effective than the electroless
system they had used previously. They reported spending less time on maintenance and lab
analysis while reducing overall cycle time. When asked to compare Envision DPS™ to electroless
copper, Facility Q stated: "I know I'm going to save a considerable amount of money." Facility
P, noting significant savings in labor and waste treatment costs, called the technology "very cost-
effective." The Envision DPS™ system also simplified waste treatment for all three
manufacturers. The facilities reported reductions in sludge generation and copper discharge,
although overall usage remained constant. Since chelating agents are not used, chelated copper
does not enter the wastestream.
Keys to Success
Each of the manufacturers offered advice for
the successful implementation of Envision
DPS™. Facility O stressed the importance of
drilling quality and consistent copper
deposition during plating. Facility Q
emphasized line control, lab analysis, and
operator training as the most important
components of success. Facility P noted that the DPS system is slightly more sensitive than
electroless to duty rinse water and to temperature. This facility found it needed to keep rinses
cleaner and to monitor temperature more closely than it had with electroless.
Facility Q emphasized that line
control, lab analysis^ and operator
,t s-j&p-Z- - - ^
training are the most important
components of success.
For more information on the Envision DPS™ system, contact Kathy Nargi-Toth of Enthone-OMI,
Inc. at 203-932-8635.
Palladium Methods
36
-------
HN504™ (Solution Technology Systems)
Background
Solution Technology Systems' s HN504™ patented process uses vanillin in the formation of its tin-
palladium colloid. Since the vanillin will attach to most surfaces except the surface of other
vanillin molecules, the vanillin on the surface of the colloids prevents them from agglomerating.
The vanillin also promotes colloid adsorption on the substrate surface, resulting in a conductive
layer of palladium. Subsequent treatment in an alkaline accelerator containing copper ions forms
a palladium-copper complex with greatly enhanced plating potential.
The HN504™ method can be run as either a non-conveyorized (vertical) or conveyorized
(horizontal) system. Approximately 70 percent of the customers using this process run multi-
layer boards to some extent. According to the vendor, the HN504™ process has plated hole
diameters as small as 0.001 inches, has run an aspect ratio as high as 21:1, and has successfully
processed Teflon®, polyimide, and FR-4 substrates. At one facility interviewed, Teflon® and
certain types of polyimide were processed twice to ensure complete void-free coverage.
Implementation at Specific Facilities
Two facilities (Facilities R and S) that have
successfully implemented the HN504™
process were interviewed for this report.
Their primary motivations for switching to the
new process included the elimination of
formaldehyde from the process, weiste
treatment simplification, and the relatively low
cost of the new system. Facility R had
originally implemented a palladium system
licensed from Solution Technology Systems. After running the system for one year and
encountering some stability problems with the bath chemistries, the facility decided to "go to the
MEASONS WHY FACILITIES
SWITCHED TO HNS04™
* BlinaaatioB of formaldehyde
»• Ease of waste treatment
*• Relatively low system cost
*• Ease of conversion
Palladium Methods
37
-------
source" of the technology and implemented the HN504™ system. This facility had already tried a
carbon system and is currently researching and testing a graphite system to complement the
existing HN504™ line. Facility S chose the HN504™ system because of the system's ease of
conversion and competitive price. At the time of installation (1990), Facility S believed that the
HN504™ process required the fewest equipment changes when converting from an electroless
copper line.
Both facilities took one to two days to install the non-conveyorized (vertical) process, with
debugging taking up to six months before the system was put into full production. Existing tanks
from an electroless copper line were used, which minimized the new equipment purchases
required. Heaters, pumps, and filters were installed in the conditioner and accelerator tanks.
Also, the conditioner tank needed a stainless steel liner, and the catalyst tank required a water
jacket for indirect heating. One facility had unexpected expenses: an electrolytic regeneration
unit was needed for the permanganate desmear bath to better control the buildup of manganese,
and a dryer was added at the end of the line to ensure that the boards were completely dry.
Solution Technology Systems also mentioned that rack agitation and a liner in the accelerator
tank to prevent acid leaching from the tank walls may be required at some facilities.
Experiences with the HN504™ Process
Neither facility encountered any problems during debugging, and both thought that the process
was very simple. There have been no customer acceptance issues for either facility. Solution
Technology Systems stated that some subtle problems can be encountered with the HNS 04™
process. For example, adding too much conditioner can result in hole wall pull-away. Currently,
one facility occasionally observes "smutting" -- an oxide film on the surface of the board — after
the flash-plate as a result of poor rinsing.
Palladium Methods
38
-------
Comparisons to Electroless Copper
Both facilities noted that they spend a lot less time on lab analysis compared to electroless copper,
since the baths are easier to analyze and need much less attention. In addition, they found that
their board failure rates were much lower with the HN504™ process. After switching to the
HN504™ process, both facilities simplified their waste treatment because there was much less
,„*„„+..«„„„, Air,« both facilities reported a decrease in sludge
copper and no chelated wastestreams. Also, x
generation and a slight decrease in water usage with the HN504™ process.
Keys to Success
Both the facilities and the vendor stressed the
importance of training for successful use of
the HN504™ system. The supplier stated that
operators need to have the desire and
willingness to make the new system work.
Both facilities feel that educating operators is
Both the facilities md Solution
Technology Systems stressed the
importance of training for successful
use of the HNSOf system.
very important. They should understand the
lab analyses and know what to look for to keep the system operating properly. The facilities also
emphasized that operators need to maintain the baths to vendor specifications. The manufacturers
noted that this system is fairly simple and does not require special equipment.
For more information on the HN504™ system, contact Eric Harnden of Solution Technology
Systems at 909-193-9493 .
Palladium Methods
-------
-------
5. CONDUCTIVE POLYMER METHOD
This process deposits a conductive polymer layer on the substrate surface of the via. A
cleaner/conditioner step coats the glass and epoxy surfaces in the through-holes with a water-
soluble organic film. A permanganate catalyst solution then deposits manganese dioxide (MnO2)
on the organic film (through oxidation). This only occurs on the film-coated glass and epoxy
surfaces. Polymerization takes place when a conductive polymer solution containing the pyrrole
monomer is applied to the surfaces coated with MnO2. The polymerization continues until all of
the MnO2 oxidant is consumed, resulting in a layer of conductive polymer (polypyrrole) that coats
the through-holes. The through-holes are then flash-plated with copper.
A typical conductive polymer process has six chemical process steps (microetch, cleaner/
conditioner, catalyst, conductive polymer, microetch, and copper flash-plate), as shown in
Figure 4.
Information is presented on the following conductive polymer method:
> Compact CP (Atotech U.S.A., Inc.)
Compact CP (Atotech U.S.A., Inc.)
Background
The Compact CP process was introduced in 1988 and is used primarily in Europe. Eleven units
are currently in fiill production, but only one unit, which is still in its trial phase, has been installed
in the U.S. Compact CP is available only as a conveyorized (horizontal) plating unit. The
volatility of the conductive polymer precludes its use in an open system, because it would deposit
a black coating on the surrounding area. Facilities using Compact CP typically produce high
Conductive Polymer Method
41
-------
Microetch
Cleaner/
Conditioner
Catalyst
Conductive
Polymer
Microetch
I
Copper
Flash-Plate
Figure 4. Typical Conductive Polymer Process Steps
Conductive Polymer Method
42
-------
volumes of FR-4 boards with four, six, or eight layers for the consumer electronic and
communications industries. According to Atotech, "the buyers of these boards are demanding
about product quality, have high technology demands, and typically require small holes."
Companies that purchase or evaluate Compact CP are usually looking for a fully integrated
system that can handle higher technology demands.
No limitations have been identified in terms of number of layers or hole size. The substrates,
however, are limited to those that react well with permanganate. FR-4 is best suited for the
Compact CP process. Polyimide also works but is not commonly used. Teflon® does not work
with this technology.
Implementation at Specific Facilities
One facility (Facility T) that is
experiencing success during the trial
phase of implementing Compact CP was
interviewed for this report. According to
the Vice President of Process and Quality
Control, the facility was motivated to
make a switch to direct metallization
because the company "looked at where
the industry was going in the next five
REASONS WHY A FACILITY
SWITCHED TO CoMFACt CP
* Faster cycle ta
* Less handling
* Lower waste treatment costs
»• Works well with small tidies
* Chemistry and equipment are available as a
package
years or so, and it was not electroless."
The company wanted a conveyorized (horizontal) process line to reduce handling and cycle time.
The company anticipated spending less money on waste treatment with the Compact CP system.
In addition, the company was concerned that the colloidal dispersions utilized in graphite and
palladium systems would not move well through the holes. The company felt that a conductive
polymer method would work well with 0.010 inch drilled (0.006 to 0.007 inch finished) holes.
Atotech U.S.A., Inc. provided the advantage of supplying the chemistry and equipment as a
package. Atotech is the only supplier in the U.S. offering a complete system.
Conductive Polymer Method
43
-------
Facility T recently opened a new facility and installed the Compact CP process in a folly
automated, conveyorized (horizontal) configuration. Acid copper plating is done in-line with
Compact CP using an equipment package developed and sold by Atotech under the trade name
Uniplate. An Atotech employee from Germany worked on-site foil-time for six weeks to assist
with the start-up phase. Installation took nearly three months to complete. At the time of the
interview, the system had been running for approximately two months. Currently, the majority of
boards running through the system are double-sided, but it is anticipated that this proportion will
change primarily to four, six, and eight-layer boards over the next six months.
Experiences with the Compact CP Process
In the vendor's experience, integrating equipment, especially equipment from different suppliers,
in an automated line with the conveyorized (horizontal) Compact CP system is often the greatest
challenge during the debugging process. In general, it often takes three to five months to get the
system running at desired levels. Autoloaders bringing boards into the system from deburring and
unloaders moving boards to in-line copper plating are computer-controlled, and it can be a
time-consuming process to integrate these positioners into the system.
Facility T experienced minor problems during installation of the system. One problem the facility
faced after start-up was the sensitivity of the system's ventilation. Vapor from one tank was
mixing with vapor from another tank within the enclosed unit. Engineers had to alter the air
balance within the system by changing the belts on the blower on the roof to reduce air flow in the
entire building. The facility also had to add and re-route plumbing, which had not been
anticipated. In addition, the facility thought rinse water after the cleaning step would not require
treatment for metals, but it did.
Comparisons to Electroless Copper
There is no direct basis of comparison to a previous system because the Compact CP system was
installed at a new facility. The following comparisons between electroless copper and Compact
CP were made based on experience at a sister facility employing electroless copper. The amount
Conductive Polymer Method
44
-------
of time spent on preventive maintenance with Compact CP is a little more demanding because of
additional cleaning requirements and other tasks such as changing filter cartridges. The process
engineer at Facility T commented that "working with permanganate tends to be pretty messy. We
have to be very conscious of cleanliness." Less time is spent on lab analysis and bath maintenance
because tank pump-outs, chemical additions, and most other bath maintenance tasks are
automated. Some analysis continues to be done manually for verification.
It is too early for Facility T to tell if there is a change in board failure rates, but they stated that
"there is no indication that defect rates are any higher (with the new system)." The cycle time is
about two-thirds faster than for the electroless line at the other facility. Whereas it takes two
hours for the electroless line alone at the other facility, the new system takes only sixty minutes
from deburring through electroplating. Water usage is greatly reduced with Compact CP. At the
other facility water is used at a rate of 30 gallons per minute (gpm). For the same level of
production, the new facility uses just 8 to 10 gpm. Facility T has not tracked copper discharge
from the Compact CP line, but it plans to do this in the near fijture. Overall, Facility T feels that
Compact CP will be very cost-effective, but it does not yet have adequate cost data to draw
definitive conclusions.
Keys to Success
"Getting customer buy-in" is key to
successfully implementing the Compact CP
system, according to Facility T. A company
considering using this system also "needs
workers to commit to being on the line for
installation, start-up, and debugging to
understand changes that have been made
along the way." This is essential because it is
not possible to simply hire someone with
experience setting up this system; conductive
"(A company] neeas-workers to
commit to being on the tine for
installation; start-up, and debugging
to understand changes that have been
made along the way,"
-Facility T
Conductive Polymer Method
45
-------
polymer technology is new in the U.S., and the learning curve is considerable. Facility T
recommends that other PWB manufacturers implementing the system "take time to become
familiar with the process, and take care of mechanical issues before running product at full
production level." Finally, start-up each day is a complex process, so a facility needs to run
enough product to make it cost-effective. For Facility T, the system "runs a lot better at 400
panels a day than 100."
Atotech stressed that management needs to make a firm decision and commit to switching to the
alternative technology. Management must support all phases of the implementation, from
installation to debugging to full production. Also, a substrate type appropriate for the Compact
CP system must be used.
For more information on the Compact CP system, contact Mike Boyle of Atotech U.S.A., Inc. at
803-817-3561.
Conductive Polymer Method
46
-------
6. OTHER ALTERNATIVE TECHNOLOGIES
Two additional alternative technologies for making holes conductive were evaluated in the overall
DfE project's Cleaner Technologies Substitutes Assessment. These were the non-formaldehyde
electroless copper process and the conductive ink process. The non-formaldehyde process is
currently in use at two facilities, one in the U.S. and one in Singapore, but neither facility was
available for an interview. The conductive ink process is currently in the developmental phase for
multi-layer applications.
Other Alternative Technologies
47
-------
-------
7. LESSONS LEARNED
No matter what the technology, some common suggestions emerged from company and vendor
experience for successfully implementing an alternative MHC technology:
»• Many facilities and vendors stressed the importance of high-quality equipment for
conveyorized (horizontal) systems.
>• Since there can be major differences between direct metallization and electroless copper
processes, line operators need to be willing to accept changes and retraining.
»> Some vendors and manufacturers emphasized that facilities should take a "whole-
process" view of the MHC technology installation. Process changes upstream and/or
downstream may be necessary to optimize the alternative MHC process.
*• Perhaps the most important factor in successfully implementing an alternative
technology is a strong commitment from management and line operators to the new
technology.
As shown in the CTSA and in these facilities' experiences, alternative technologies are
successfully "making holes conductive." According to the manufacturers interviewed for this
report, alternative technologies offer benefits, but facilities may first have to overcome the
problems encountered during installation.2 After installing these systems, the most successful
facilities improved their production efficiency and their worker safety, while decreasing
environmental impacts. Hopefully, the experiences of these manufacturers will help others
considering a switch to an alternative MHC technology.
For a description of the experiences of PWB manufacturers in northern California with direct metallization,
see the report, Direct Metallization Report, completed by the City of San Jos6's Environmental Services Department.
For a copy, contact John Mukhar at 408-945-3036.
Lessons Learned
49
-------
-------
APPENDIX A. FACILITY INFORMATION
Facility Surface square feet
produced annually
Types of PWBs High
produced aspec
norma
est Conveyorized or
t ratio non-conveyorized
lly run system
Carbon Method: Blackhole® (MacDermid, Inc.)
Facility A 750,000 ssf/yr
Facility B 420,000 ssf/yr
15% double-sided 8:
85% multi-layer
30% double-sided 10
70% multi-layer
1 conveyorized
: 1 conveyorized
Graphite Method : Graphite 2000™ (Shipley Company)
Facility C 1,000,000 ssf/yr
Facility D 600,000 ssf7yr
65% double-sided 8:
35% multi-layer
2% single-sided 3:
70% double-sided
27% multi-layer
1% flex
1 conveyorized
1 conveyorized
Graphite Method: Shadow™ (Electrochemical, Inc.)
Facility E 3 50,000 ssffyr
Facility F 800,000 ssf7yr
Facility G 360,000 ssffyr
70-75% double-sided 12
25-30% multi-layer
1 0- 1 5% double-sided 8 :
85-90% multi-layer
70% double-sided
30% multi-layer
: 1 conveyorized
1 conveyorized
see below3
a Facility G switched from the Shadow™ process back to electroless copper.
Appendix A
-------
Facility
Surface square feet
produced annually
Types of PWBs High
produced aspec
norma
est Conveyorized or
t ratio non-conveyorized
Hy run system
Palladium Method: Neopact (Atotech U.S.A., Inc.)
Facility H
Facility I
26,000 ssf/yr
250,000 ssffyr
70% multi-layer 9
10% flex
10% rigid flex
10% microwave
(Teflon®)
40% single-sided
40% double-sided
20% multi-layer
1 non-conveyorized
non-conveyorized
Palladium Method: Conduction DP (LeaRonal Inc.)
Facility J
Facility K
520,000 ssffyr
700,000 ssFyr
5% double-sided 9:
95% multi-layer
99.9% multi-layer 10
1 non-conveyorized
: 1 conveyorizedb
b Facility K is currently running electroless copper while installing a new Conductron DP-H line.
Palladium Method: Crimson 1® (Shipley Company)
Facility L
Facility M
Facility N
9,600,000 ssffyr
1,500,000 sstfyr
1,440,000 ssfyr
30% single-sided 6:
60% double-sided
10% multi-layer
25% single-sided 6:
75% double-sided
33% multi-layer 6:
33% double blind via
33% PCMCIA
1 conveyorized
1 conveyorized
1 conveyorized
Appendix A
-------
Facility Surface square feet
produced annually
Types of PWBs
produced
Highest
aspect ratio
normally run
Conveyorized or
non-conveyorized
system
Palladium Method: Envision DPS™ (Enthone-OMI, Inc.)
Facility O 600,000 ssffyr
Facility P 250,000 ssffyr
Facility Q 240,000 ssfyr
10% single-sided
70% double-sided
20% multi-layer
5% single-sided
80% double-sided
15% multi-layer
20% double-sided
80% multi-layer
4:1
7:1
non-conveyorized
non-conveyorized
non-conveyorized
Palladium Method: HN504™ (Solution Technology Systems)
Facility R 36,000 ssfi'yr
Facility S 300,000 ssffyr
Conductive Polymer Method:
Facility T 960,000 ssf/yr0
10% double-sided
90% multi-layer
5% single-sided
35% double-sided
60% multi-layer
11:1
7:1
non-conveyorized
non-conveyorized
Compact CP (Atotech U.S.A., Inc.)
80% double-sided4
20% multi-layer
5.5:le
conveyorized
"Facility T's projected production.
dFacility T predicts a transition to 30-40% double-sided, 60-70% multi-layer.
"Facility T, predicts an increase to a 7 : 1 aspect ratio.
Appendix A
-------
------- |